GEO600: building a gravitational wave observatory

1
GEO600 building a gravitational wave observatory

• Martin Hewitson
• AEI, Hannover
• for the GEO team

2
International network of ground-based detectors
GEO600
VIRGO
LIGO detectors
AURIGA, EXPLORER NAUTILUS
TAMA300
ALLEGRO
NIOBE
bar detectors
Bar detectors
Interferometers
interferometers
3
GEO600 Hannover, GERMANY
4
GEO600 optical layout
North arm
4m
East arm
1.5kW
1.6W
12W laser system
Attenuate to 5W
Power-recycling mirror
600m
Signal-recycling mirror
35mW
5
Control room
1.5kW _at_ BS
MCN
BS
MCE
PR (0.09)
SR (2)
1.6W
output detector
MC 1
5W
MC 2
6
Design sensitivity

• Signal-recycling cavity tuned to be resonant at
  1kHz

7
Test-mass Suspensions
8
Locking scheme

• Power-recycling
• Michelson
• Signal-recycling

9
Measured optical transfer functions
V
dL
Amplitude (A.U.)
Phase (deg)
Frequency (Hz)
10
Data flow

• 64 high fs
• 64 mid fs
• gt1000 low fs

11
Calibration I
Michelson servo model

• Correct for Michelson servo and optical gain
  response

TF Strain ? detector output
x
Optical cavity model

12
Calibration II
CAL
P and Q
actuator
optical
dL
V
V
Use IIR filters to calibrate detector outputs to
two strain channels
13
Calibration III

• Maximum likelihood estinmator for h(t)
• Use noise-floor estimates to get weights
• Convert weights to filters

14
On-line monitoring and remote operation

• LabVIEW system guides analog loops
• Low-latency data viewing systems

15
Commissioning

• Implementing subsystems
• Mostly done now
• New laser still needed
• Stability
• Lock-loss investigations, alignment control
• Sensitivity
• Noise hunting
• Feedback systems, light and modulation noise,
  other
• Use noise-projections, intuition
• Data quality
• Stationarity, glitch rate, lines

16
Noise-projections

• Measure TF from source to detector output
• Measure PSD at source
• Multiply to project source to main output

17
Sensitivity progress
hrms/?Hz
Frequency (Hz)
18
S4 Science run - statistics

• 4 Week science run in coincidence with LIGO
  detectors
• 708 hours of science data
• Duty cycle 95
• Longest lock 50 hours

19
The way forward

• Tune to lower signal-recycling resonance (300Hz)
• Noise hunting focusing at the lower frequencies
  (100-1000 Hz)
• Feedback noise, laser amplitude noise
• Turn up laser power (?5)
• New laser setup
• Continue to work towards design sensitivity
• S5 extended science run (gt1 year)
• Starting end 2005